5 CIO priorities that won’t wait for AI in 2027

TechTarget

Source published: Site published:

Original article · فارسی

A business professional stands in a meeting room with colleagues
In 2027, CIOs face many technology and business priorities that go beyond AI.Getty Images

AI may be the biggest issue for CIOs in 2027. It will affect decisions about business software, technology systems, security, budgets and how people work. But AI adds to existing financial, technical and organizational challenges. It does not replace them.

And decisions are not always settled when a project ends or a contract is signed. A system migration might finish on time but fail to save as much money as planned. A contract might promise that a company can leave a supplier, even when there is no realistic alternative.

CIOs are still responsible for business results, even when they do not have full control over the technology and suppliers involved.

1. Technology costs become harder to predict

As 2027 approaches, CIOs face a difficult situation: Technology spending is growing quickly, but the budgets available to individual IT departments are not keeping up with some of their rising costs.

Gartner's 2027 CIO survey predicts that IT budgets will grow by an average of 3.7%. A separate Gartner forecast predicts that worldwide technology spending will reach $7.35 trillion in 2027, a 15% increase from 2026. These figures measure different things: worldwide technology spending and the budgets of individual CIOs. They cannot be compared directly. Together, however, they show the different financial pressures organizations face.

Hardware costs are making the problem worse. Gartner says memory prices more than doubled during the previous year. This has increased the cost of servers, storage systems, networking equipment and PCs. Gartner expects these higher costs to continue putting pressure on budgets in 2027. Demand for AI infrastructure is one reason for the increases, but companies buying ordinary IT equipment must also pay more.

Marc Hoit, CIO at North Carolina State University, told InformationWeek that server price quotes were two or three times higher than they had been a month earlier. His team was considering using hardware for longer, reusing memory and buying less equipment.

IT financial management is also covering more areas. The FinOps Foundation's 2026 survey included 1,192 people who work in the field. Of the 632 people who answered a question about which costs they managed, 90% said they were managing or planning to manage software-as-a-service (SaaS) spending. Another 64% said the same about software licenses. The same group reported figures of 57% for private cloud and 48% for data centers.

These findings show why cost decisions now involve more than cloud bills. They also include software contracts, how long infrastructure is used and where companies invest their money.

Combining software products does not always lead to easy savings. Futurum analyst Keith Kirkpatrick told TechTarget, “Buyers are planning consolidation faster than they are funding it.”

Infrastructure contracts can involve similar decisions. At Danish manufacturer Danfoss, changes in supplier costs helped lead to a plan to move more than 2,500 virtual machines away from VMware. The company aims to finish the move by the end of 2027. It also wants a simpler and more consistent technology setup across its business.

“It’s much more to the architectural thinking of our infrastructure, where we talk about a digital backbone across the company,” Danfoss CIO Sune Tornbo Baastrup told TechTarget's Paul Crocetti.

The project covers about 1,100 servers, so the move itself requires a major investment. Whether it saves money will depend on the costs and features of the new system after the move is complete.

The challenge is not just knowing what a technology platform costs today. CIOs must compare the cost of staying with the cost of leaving. That includes the work of moving systems, the risks of continuing to maintain old technology and the new dependencies that either choice might create.


“Buyers are planning consolidation faster than they are funding it.”

Keith Kirkpatrick

Vice president and research director, enterprise software and digital workflows, Futurum


2. Modernizing old systems runs into real deadlines

Technology suppliers do not always set their support deadlines to match an organization's plans to replace old systems. This difference will become especially important in 2027. Regular extended support for Windows Server 2016 ends in January. Support for SQL Server 2017 ends in October. Regular maintenance for covered SAP Business Suite 7 applications ends in December.

Organizations that still use these systems must decide how much time they need. Depending on the product, they might be able to buy extra support or security coverage, but this costs money. Replacing a system before the business is ready can also be expensive.

Microsoft's own move of a 70-terabyte SAP billing system shows how difficult this work can be. An early review suggested that the system would need to be unavailable for five days. That was too long for a system used to handle transactions around the world.

“We were not going to be able to do the project unless we could fit it into a weekend,” principal SAP solution architect Melinda van Honschooten told TechTarget's David Essex.

The project began in 2024, and Microsoft brought together its full project team in February 2025. The team moved selected data, carried out extensive tests and carefully planned the changeover. These steps reduced the outage to about 24 hours. But the work was not over. Another major ECC system used for finance and human resources still needed a separate migration that would take several years.

At manufacturer John Boos & Co., the problems were different. When Essex spoke with company executives, a move to SAP S/4HANA Cloud, planned for April 2027, was already behind schedule. The project uncovered more than 120 custom programs, processes that had never been properly documented and poor-quality data.

Moving to a more standard platform meant rethinking ways of working that employees had followed for years. “Our data is garbage,” business development vice president Ty Jones told Essex. “Data hygiene was never part of the culture of this business.”

By the numbers

Challenges in modernizing SAP

A 2026 survey of people who work with SAP found that many organizations had made progress with migration but still faced problems with custom software and data quality.

67%
Said their organizations had moved fully or partly to SAP S/4HANA. A partial move does not mean the changeover is complete.
53%
Said dealing with customizations in SAP ECC was one of their main migration challenges.
83%
Said cleaning up SAP master data was important for the migration.

Source: Precisely/ASUG, 2027 Trends and Challenges (2026 survey of SAP stakeholders).

In ISG's 2026 research on SAP migrations, 58% of the programs studied missed both their budget and their schedule. Another 27% missed either the budget or the schedule. Only 15% met both targets. These results describe the programs included in the research, not every SAP migration. They show how changes in project size, poor management and the need to run old and new systems at the same time can make modernization harder.

A project might be technically ready before the business, its data or its work processes are ready.

3. Security risks from suppliers become a company-wide responsibility

A security problem does not have to begin inside a company to cause trouble for it. Verizon's 2026 Data Breach Investigations Report found that third parties were involved in 48% of the security breaches it studied. This does not mean that suppliers caused all these incidents. But it shows why CIOs must think about risks involving systems and companies they do not directly control.

The World Economic Forum's 2026 cybersecurity research suggests that companies may be getting better at checking suppliers than at understanding how much they depend on them. Among survey respondents, 66% checked suppliers' security readiness, and 65% included security teams in purchasing decisions. Far fewer, 33%, said they had fully mapped their supply-chain networks. Only 27% said they carried out cyberattack practice exercises or recovery tests.

These figures describe what organizations say they do. They are not independent proof that the organizations are ready for an attack. But they point to an important difference. A company might know that a supplier passed a security check without knowing which applications or business activities would stop if that supplier suddenly became unavailable.

A supplier that seems unimportant might still have access to sensitive systems or provide software that is part of a critical application. Ranking suppliers helps companies decide where to focus security work. But those rankings may not show every way a problem at one supplier could affect the business.

In October, Lora Vaughn, founder of Vaughn Cyber Group and a CISO, described how community banks were struggling to identify vulnerable software in their own systems and in systems managed by suppliers. “Plenty of teams couldn’t answer that in a day,” she told TechTarget's Sharon Shea.

“CIOs have to realize they’re running the biggest vendor management department in the company,” Swingtide CEO Diane Carco told Mary K. Pratt.

Another issue for 2027 is the EU Cyber Resilience Act. Its main product-security rules begin to apply in December to manufacturers and other businesses covered by the law. Reporting requirements began in September 2026.

The main product-security rules do not set one universal compliance deadline for CIOs who buy these products. However, they give CIOs another reason to check how suppliers deal with security weaknesses, provide security support and meet their contract promises throughout a product's life.

Using fewer suppliers might make them easier to manage, but it can also mean that more important business activities depend on a smaller number of providers. Contracts and security checks help make responsibilities clear. But they cannot replace a clear map of the business processes and recovery plans that depend on each supplier.


“CIOs have to realize they’re running the biggest vendor management department in the company.”

Diane Carco

President and CEO, Swingtide


4. Keeping operations running takes more than preventing outages

A system that works reliably is not always a system that can recover quickly when something goes wrong.

Outages are becoming less common, but they can still cause major financial losses. Uptime Institute reported in 2026 that the number of outages at each site had fallen for five years in a row. However, in its 2025 survey, 57% of respondents estimated that their most recent major outage had cost more than $100,000. One in five said the damage was more than $1 million.

CIOs also need to think about another problem as they prepare for 2027. Uptime's research points to more reported problems involving fiber-optic cables and network connections. Yet companies often focus their recovery checks on the systems they operate themselves. This means some important connections and dependencies may be missing from their recovery plans.

This problem became clear in October 2026, when a power failure in one availability zone in Google Cloud's London region disrupted banks, delivery companies and government services. Some customers had to keep their workloads in the U.K. because of their location requirements. They faced a problem because Google had no other cloud region in the country that they could use as a backup.

Computer Weekly reported that most affected services recovered that afternoon. But the Ministry of Justice said its court video system had not fully recovered. The cloud provider getting its systems working again did not necessarily mean that its customers had returned to normal operations.

Having backups does not solve every recovery problem, either.

“What’s commonly missing from recovery plans is a clear map of service dependencies and the operational steps required to re-establish them in the right order,” former Forrester analyst Brent Ellis told Sean Michael Kerner.

The 2024 UniSuper incident is one example. The Australian financial institution lost its Google Cloud subscription because it was accidentally deleted. Ellis mentioned the incident in his discussion with Kerner. Backups were available, but restoring the data alone was not enough to get services running again. The company also had to rebuild the surrounding environment, including its networks, identity systems, settings and connections between systems.

By the numbers

Restoring technology is not the same as keeping a business running

IT and security decision-makers who answered a survey reported a gap between restoring computer systems and keeping essential business activities going.

78%
Said their organizations focused more on restoring systems after a cyber incident than on keeping business operations running.
22%
Said their organizations had officially written down and tested a plan for a “minimum viable company” — the essential activities that must continue during recovery.

Source: Cohesity, 2026 Global Cyber Resilience Report (Vanson Bourne survey).

This leaves CIOs with difficult decisions about how much money to spend on backup systems, recovery tests and alternative infrastructure. These protections cost money, but so does a long interruption to business operations.

A backup or service-level agreement is a starting point. When something fails, the important questions are whether the organization knows what must be restored, what order to restore it in and who can do the work.

5. Cloud control and the ability to move become practical buying questions

Cloud platforms give CIOs access to technology and services that would be difficult for them to build themselves. But using these platforms also creates dependencies. If a supplier changes its prices, rules or services, how easily can a customer choose another provider?

This question becomes more important in 2027. Starting January 12, the EU Data Act requires covered data-processing providers to remove certain charges for switching providers. These include related fees for moving data out of a service. This is an important change in the cost of moving workloads, but it does not make every type of data transfer free. It also does not remove the work needed to move applications, rebuild connections between systems or keep business operations running during the change.

The difficulty is shown in a 2026 survey commissioned by Leaseweb. It questioned 800 senior executives at companies with 500 to 5,000 employees in four countries. About 89% estimated that they could not move important workloads away from their largest cloud or infrastructure provider within three weeks.

The 89% figure is based on executives' estimates, not actual tests of how long a move would take. Three weeks would also not be a realistic target for every workload. Still, the answers suggest that many companies believe they have little freedom to change providers. Getting data out of a cloud service is one task. Moving the applications and connections built around that data is another.

By the numbers

Control over cloud data does not guarantee that workloads can be moved

A survey of executives asked about their ability to move data and how long they thought it would take to move important workloads.

76%
Said they had complete or strong control over their ability to move data to another provider or environment.
55%
Estimated that moving important workloads away from their largest cloud or infrastructure provider would take more than three months or was not currently possible.

Source: Leaseweb, Digital Independence Survey 2026 (800 executives; estimates, not tested migrations).

Some organizations are trying to gain more direct control over their technology. For example, the Dutch government is developing a sovereign cloud design based on open source software.

The Dutch government does not claim that it can operate without outside suppliers. Officials accept that the proposed design still depends on hardware made by companies outside Europe. Open source software could give the government more control over some parts of its infrastructure, but it would not remove all outside dependencies. The design is still a proposal, not a completed move.

Other organizations are making different choices. Nationwide Building Society renewed its VMware Cloud Foundation agreement with Broadcom. It said it needed the platform's reliability and ability to work with other systems after buying Virgin Money. This decision came at a time when other VMware customers were thinking about leaving the supplier.

These different choices do not necessarily conflict. The Dutch government wants more independence, while Nationwide has reasons to keep using a platform it already knows. Both must decide what they need from suppliers and how much it would cost to change providers.

“The big trick is architecting in a way that fully uses the platform while avoiding irreversible decisions,” Southern Glazer's CIO Steve Bronson told TechTarget's Tim Murphy.

Forrester analyst Dario Maisto has recommended an approach called minimum viable sovereignty. It means balancing legal rules, business needs and the practical limits of becoming independent from technology suppliers.

Not every application needs to be easy to move, and reducing dependence on suppliers is not always worth the cost. For some workloads, keeping another option available may be worth the extra spending and technical work. For others, a closely connected platform may provide benefits that are greater than the risks.

CIOs need to understand where their organizations have the freedom to make changes, where they do not and whether they could realistically change direction if they had to.


“The big trick is architecting in a way that fully uses the platform while avoiding irreversible decisions.”

Steve Bronson

CIO, Southern Glazer’s Beverage Company


The decisions still need to be made

Finishing a system migration is one thing. Keeping the new system running, secure and ready for recovery is another. Supplier contracts bring their own problems. An agreement that makes sense today might look very different if a supplier changes its prices, products or plans.

AI adds new demands to these long-standing responsibilities and makes the connections between them even more important. But business technology decisions have never been simple or straightforward.

CIOs will make separate decisions about budgets, modernizing systems, security and suppliers. Their businesses will have to live with the combined results, including the benefits, difficult choices and unexpected problems that follow.

James Alan Miller is an experienced technology editor and writer and Lead Editor for CIO News at Informa TechTarget. He leads coverage of business technology strategy, AI, software, data, infrastructure and the decisions that shape how CIOs manage increasingly complex IT systems.